Why verbalization of facial features increases false positive responses on visually-similar distractors: A computational exploration of verbal overshadowing.